<  Back to the Polytechnique Montréal portal

Bispectrum and recurrent neural networks: Improved classification of interictal and preictal states

Laura Gagliano, Elie Bou Assi, Dang K. Nguyen and Mohamad Sawan

Article (2019)

Published Version
Terms of Use: Creative Commons Attribution .
Download (870kB)
Cite this document: Gagliano, L., Bou Assi, E., Nguyen, D. K. & Sawan, M. (2019). Bispectrum and recurrent neural networks: Improved classification of interictal and preictal states. Scientific Reports, 9. doi:10.1038/s41598-019-52152-2
Show abstract Hide abstract


This work proposes a novel approach for the classification of interictal and preictal brain states based on bispectrum analysis and recurrent Long Short-Term Memory (LSTM) neural networks. Two features were first extracted from bilateral intracranial electroencephalography (iEEG) recordings of dogs with naturally occurring focal epilepsy. Single-layer LSTM networks were trained to classify 5-min long feature vectors as preictal or interictal. Classification performances were compared to previous work involving multilayer perceptron networks and higher-order spectral (HOS) features on the same dataset. The proposed LSTM network proved superior to the multilayer perceptron network and achieved an average classification accuracy of 86.29% on held-out data. Results imply the possibility of forecasting epileptic seizures using recurrent neural networks, with minimal feature extraction.

Uncontrolled Keywords

biomedical engineering, epilepsy

Open Access document in PolyPublie
Subjects: 1900 Génie biomédical > 1900 Génie biomédical
1900 Génie biomédical > 1901 Technologie biomédicale
Department: Département de génie électrique
Institut de génie biomédical
Research Center: Autre
Funders: NSERC / CRSNG, Epilepsy Canada, Institute for Data Valorization (IVADO)
Date Deposited: 14 Jul 2021 11:19
Last Modified: 15 Jul 2021 01:20
PolyPublie URL: https://publications.polymtl.ca/4876/
Document issued by the official publisher
Journal Title: Scientific Reports (vol. 9)
Publisher: Nature
Official URL: https://doi.org/10.1038/s41598-019-52152-2


Total downloads

Downloads per month in the last year

Origin of downloads


Repository Staff Only